Uber has cut 10 percent of its customer-service operation. The company says the move supports a wider restructuring. It also says the plan will “embrace artificial intelligence”.
The decision gives the contact-centre market a clearer signal. AI no longer only supports agents during calls. It now handles many conversations before humans join them.
For years, vendors framed automation as an agent-assist tool. It wrote summaries, suggested answers, and searched knowledge bases. Now, AI chat and voice tools resolve routine tasks alone. That changes the workforce equation.
Ashish Nagar, founder and CEO of Level AI, expects customer-service employment to shrink. He told UC Today that job numbers may fall 10 to 30 percent. He expects that shift over five to 10 years.
“Once a week I get the question, what happens to CS [contact centre] jobs in three to 10 years?” Nagar said.
“My answer has been: a 10–30 percent decline from current levels over five to 10 years and dependent on verticals.”
That does not suggest a sudden collapse. However, it challenges a common industry promise. AI may improve productivity, but it can also reduce headcount.
In Uber’s case, the company linked the restructuring to operational change. A spokesperson said it aimed to “simplify operations, strengthen in-person collaboration, and continue to embrace AI”.
Megha Yethadka, Uber’s vice president of global community operations, also told staff: “We cannot scale frontier technology on top of fragmented processes.”
Other large employers show similar patterns. Commonwealth Bank of Australia reportedly reduced contractor roles in Johannesburg. Its AI chat platform resolved nearly nine in 10 conversations without agents.
Microsoft has also expanded support automation. Reports suggest its support workforce fell from about 50,000 to 40,000.
The business case remains simple. AI can handle basic requests at any hour. It can manage password resets, order tracking, booking changes, and receipt requests. These tasks once filled large agent queues.
Hyatt Hotels uses AI for reservation changes and receipt requests. Pat Nestor, who leads its AI and data analytics work, said cost reduction was “clearly a driver”.
Still, replacement will not happen everywhere at the same speed. Customers often want humans for sensitive issues. Healthcare, insurance, and banking require more trust.
Nagar also pointed to integration barriers. “It’s the API, not the AI,” Nagar said. In simple terms, AI must connect with older business systems. Without that access, it cannot complete many transactions.
He added: “The easy stuff was already automated; new capabilities just do it better.”
For telecom and VoIP teams, the message is clear. Contact-centre design now needs both automation and human escalation. Companies will chase lower costs and faster service. Yet they must protect customer trust, especially during complex moments.
AI will reshape support gradually. But the direction now looks much harder to deny.
